In trying to tune/debug my idle, I've been seeing my system go into a mode I can't understand. I mostly see it at idle with the vehicle in drive (brake held). What it's doing is showing consistently high O2 readings (like cycling from 145 to 170's) and then also showing very large BLM values (160+). It's like it's stuck rich and isn't even trying to fix it.

What might cause the ECM to give up even trying to get the O2 right, or else to purposely target a rich O2 (unlikely)? And why would the BLM's be so big under these circumstances? It really seems like the ECM is doing this to itself on purpose, because when it comes up open loop, the O2 readings are off the chart low (an entirely different problem that I'm working on).

 

If the O2 readings (cycling from 145 to 170's) are in milli-volts then those are low. That is why the ECM is at 160+ BLM.

This can be caused by a one-wire O2 too far downstream (cooling off), by an exhaust leak, or an improper tune.

RBob.

 

No, those are converted #'s. The mV's are 0.6 to 0.74 or so.

Does the answer change now?

 

Oh, I do have a 3-wire heated O2 by the way.

I checked like crazy for an exhaust leak but could find none except at the collector, which is WAY downstream of the O2. Is there any way that could muck things up? The leak isn't so bad, but I can tell it's there.

 

Is teh O2 cycling between around 200 mV and somewhere around 800 mV?

 

No, that's why I don't understand it. It cycles around 600 to 750 mV. The ECM seems like it's not even trying to make it leaner, and the BLM values are huge.

I'm hoping all my problems are somehow O2 sensor related. I have a Bosch, and after all the bad stuff I've read here, I ordered a Delco one. Got my fingers crossed.

 

Here' some WinALDL data in case my verbal description wasn't clear:

 

I think the 7747 has memory with it's BLMs on restart. You might want to try unplugging the ecm (or battery) and resetting everything JUST to be certain that isn't an issue. I'm just assuming a freak situation where the idle cell was jammed somehow from an overly lean idle and now that it's fixed it's freaking out. BTW, those o2 values really aren't all that rich. Usually above 850 constant is RICH.
I really can't think of what's going on without being there in person. I'd just try different things and take notes.

edit: It just occured to me, what is your 0rpm VE in the VE adder table? If that's 0 you can screw up a low engine speed idle.

 

That data is definatly odd. Is it possible for the ECM to, well, get stuck if it doesnt see the correct output from the O2 (ie malfunction somewhere with the O2)? I think if it hangs outside the rich or lean bounds too long it exits closed loop but maybe if it stays around 'stoich'?

 

Does the engine run OK? Are the plugs carboned up?

This location in the '747 adds to the O2 window value when in idle. This will cause the O2 to swing at higher levels:

LD2F7: FCB 25 ; 200 mvdc o2 SENSOR BIAS AT IDLE

If the engine runs OK it may just be that the VE table and the above parameter need to be changed.

RBob.

 

I'm pretty certain that I reset the ECM every time I tried a new chip. Hmm, where is the BLM stuff stored anyway? Can it persist when the PROM is reprogrammed?

My 0 RPM VE table 2 entry is 25%, so that should be okay I think.

 

Occasionally it went out of closed loop, and I figured this was because it blew the AFR limits set for closed loop operation.

One thing I don't understand is under what circumstances the system normally goes to open loop idle. There are parameters in the bin called "open loop idle on - RPM" and "open loop idle off - RPM". I've looked at some different bins, all supposedly factory stock, and some have 850 and 750, others have 50 and 25. What exactly do these parameters do? At present, my system never seems to go into open loop at idle unless its cold or if something isn't working right with closed loop.

 

Yeah, I have an entry that tunerpro calls "O2 idle AFR bias value". It's set to 109 mv. So does that mean that instead of targeting 450 mv, the ECM would try to bracket 559 mv? Even so, what I'm seeing is high.

Actually though, yeah, the system is running okay. The no load idle (in park) surges a little, like +/- 50 RPM, but it's not that bad. I haven't pulled a plug, but I pulled the O2 sensor last night and it looked clean. I expected it maybe to be black.

I guess if I didn't have the computer hooked up I wouldn't be too unhappy. Seeing 160+ BLM's though kinda freaks me out and I'd like to know what's going on. I can't seem to throw enough fuel at it in idle to get the BLM's down - then I get into this weird mode where it really does look rich, but the BLM's are still high anyway. I'm at a loss.

No matter how big I make my idle cell VE, the engine always fires up with an O2 read of 108# (not mv), then after 20-30 seconds drops down down down to 0. It stays at 0 until closed loop kicks in.

Is there any limit to how much the ECM will pull away from the base VE table in closed loop? The only thing I see for sure is the BLM limits, which I'm not getting to (108/172).

Thanks for the help.

 

Do you have AIR? On mine its usually around 450 mV untill its warm enough to start working.

 

Yes, those blm values detirmine how much the ecm can adjust the fueling. Dont want it to have too much adjustability in case there is a problem that skews the O2 readings.

Are you using a stock chip? If not give stock calibration similar to your setup a try and see if its any better. The bias should cause it to swing around the higher value at idle. Yours looks kinda odd but maybe getting the fuel in line would make things better.

 

It is at 450 mV, but the number in winALDL is 108 or so, not 128, which I think is the "target" O2 number, right? Darn, I'll have to go look again because now I'm getting confused. Maybe the 109 mV bias I have makes 450 mV into 108, and if my bias was zero then 450 mV would be mapped to 128?

No, I don't have the air pump stuff.

 

Just look at the mV for now. I never was sure how the R/L indicator was supposed to work in winALDL. Maybe its sort of an average of when its rich and lean?

 

Maybe I told you guys the wrong thing. I could have sworn the data above was in closed loop, but I just ran the experiment again and it was not in closed loop. Well, it was in closed loop in park, then as soon as I went to drive it went open loop.

Can someone explain to me how that works? How does the ECM decide when to go into open loop idle, and why would you want it to?

Thank for the help.

 

Thats wierd, sounds like your using the wrong polarity P/N switch. Jsut pull the wire off the switch and make the ecm think its in drive all the time.

 

Yeah, I don't know what that is either. What I was looking at was the O2 on the "sensors" page. It has mV, but also just raw numbers. I always thought 128 was what I wanted on average. You're probably right though, best just to look at the mV.

 

The raw data? Thats the raw unformatted output from the ECM. Wont mean much unless its converted. Ignore it.

 

Mastiff, everything I am reading in your posts make sense. From what I can tell the engines idle fueling is doing exactly what it should be. The 108 reported for the first 20-30 seconds is due to the O2 sensor not being ready (not up to temperature).

Then until closed loop the O2 sensor reports a low reading. Once closed loop is enabled the BLM moves up until the O2 sensor is swinging.

Now, it may seem as though the swing voltage is high, but I don't believe it to be. The idle O2 bias is added to whatever is in the O2 term vs airflow tables.

There is also an O2 correction term for whenever in idle:

LD2FC: FCB 192 ; 0.750 IDLE ERR CORR TO ERROR

This actually reduces the error and allows a greater deviation in the desired O2 swing.

You have pulled the O2 sensor and it is clean. Can't hurt to pull a plug or two. Just do it after idleing for 20 minutes or so (and an engine off cooling).

Maybe the VE table. Double check the log file for idle RPM & KPa. If the BLM is still at 160 then do this:

NewVE = CurrentVE * 160 / 128

Plug that into the VE table and see what happens. May need to do various cells in and around the idle RPM/KPa.

RBob.

 

I don't have that even hooked up. My only guess is that it's working off of RPM somehow. The RPM goes down a little in park, maybe it is going below the threshold for open loop idle.

I bet that's it. One thing's for sure, the idle is much smoother in open loop. Is that why they have open loop idle, to make it idle smoother instead of oscillating around a little chasing the O2?

 

LOL, you guys are having a conversation while I'm typing

As long as the Prk/Nt line is open that is OK. Open is in drive.

Interesting about the open loop idle being smoother. Another item, if the BLM of 160 has been learned in (or actually whatever it is) that is still being used while in open loop.

In this case the real difference is that the proportional gains are zero'd while in open loop.

What is the engine configuration?

RBob.

 

My engine is a mild 350 with headers, no EGR, no cat, no AIR, and open element air filter. The situation is that I'm converting from carb and I've never had a really good running chip. I worked through customefis.com since I knew nothing when I first started this project. My best working PROM is from customefis, but he won't send me the bin because of some warranty issue or something. I haven't read it because it's the only PROM that works well and I'm paranoid about destroying it getting it out of the carrier.

What I do have is the first chip he sent me (the other is the second, with only the VE turned up a little I think).

I've been trying to start from a factory 350 bin and work into a good tune, but they are incredibly lean at idle no matter what. I've turned up the VE so much that I'm about to pass out while the car is running, but the O2 still isn't happy. The only idea I have is that it's related to the O2 109mV bias which is on one chip and not the other.

I'm hoping some of these problems will just go away once I replace the Bosch O2 with the Delco, but I'm not holding my breath.

 

Replacing the eprom doesn't reset the BLM learn cells. Definatly reset the ecm power and retry an older eprom, maybe an original GM calibration.

 

I take it that the exaust smells real rich at idle? That sounds like a false lean condition.

JP, thats interesting about it retaining the BLMs. I thought the data in the ram area was lost on shutdown. Guess its not.

 

You know, I'm not so sure anymore. I thought it smelled stinky as heck, but I tried it again last night. My BLM's at idle were like 160+, so I started cranking on the VE to see what it would take. I basically doubled the VE in my idle cell (800,30) from 17% to 30% and it barely moved, maybe down to 150-155. I can't explain that.

I decided to experiment by going after the BPW instead to see what would happen. Long story short, I wound up cranking it to 175 (135 was stock) and that brought the BLM at idle exactly to 128. The engine seems happy like that. I don't really understand how this could be. It's possible that it actually is very rich, but not enough for me to notice. Or possibly the O2 is telling the truth and something else is weird that I am compensating for with an extra large BPC.

The only theories I have besides bad O2 readings (maybe due to exhaust leaks?) are that my injectors are small, or that the ECM thinks it's commanding EGR and is leaning accordingly (I have no EGR).

I have the EGR enable VSS up at 255, so I don't think it should be coming on. I could possibly just probe the EGR pin on the ECM to see what's happening, but I'm not sure whether voltage is applied to enable EGR or disable it? I could also try changing the EGR VE compensation table to all 1's, right?

As far as injectors go, is it possible to identify what I have without disassembling the TBI? I'm not sure where to look for the color codes.

 

It sounds like it jsut doesnt have enough fuel then. After youve determined whay your BPW roughly should be, add fuel into the main VE table till it smooths out and see what the blms are. If 128 still seems a bit flakey then add in more of the afformentioned O2 bias to help smooth the idle. Keep in mind that there is a VE adder table as well so the changes you make in the main VE table arnt as large as you think as there is additional VE from the adder table that is added in with the main VE to get the overall VE. Also make sure to add around the areas taht it idles in as it may wander from cell to cell and sudden jumps will effect the motors operation. Its also important to remember that the computer interpolates between cells.

Either on the posts or on the metal thingy between them there should be dabs of paint. Get those colors and look up the injector. Also, dont forget to make sure your running the proper stock pressure, or take the fuel pressure into accont if your running higher, when your calculating the BPWc.

 

I see what you are saying. When I started, my FL1 entry was 17% and FL2 was 39%. So When I went up to 30% in FL1, I didn't increase it by 76% (30/17), like we are often led to believe, but by (30+39)/(17+39) = 23%.

In any case though, I boosted the VE by 23%, then boosted the BPW (175/135) = 30%, so something is way off from the factory calibration I started with. It looks like I increased the fuel by 60% (1.23*1.3). If I tried to do that with VE alone, leaving my BPW at 135, it'd take a 90% VE at idle. I assume that's really not possible.

So what the heck? If my injectors are small, that explains everything. Otherwise, I'm at a loss.

Here's what I can see. I don't see any paint. Is it where the rust now is? All I see is that the posts themselves are orange. Thanks for all the help.

 

There are three choices: injectors too small, low fuel pressure, nice large displacement. I've modded your picture to show where the paint dabs woudl be. I even used white/green as with stock 305 injectors

RBob.

 

Thanks Rbob. I'll look REALLY close when I get home and see if there are a few flecks of paint left...

Well, my bin is for a 5.7L (350), so if my injectors happen to be 305, then based on the chart I found on another thread, I'd be off by ~20% for that (55 vs. 45 #/hr).

What does the factory expect for fuel pressure, any idea? Because I'm typically getting around 10 PSI. I'm slightly concerned because the pressure changes a little, sometimes it's as high as 12, sometimes as low as 9. It generally seems to get lower as the engine warms up, which kind of makes sense, since the regulator is just a spring, right? It'd probably compress more easily when hot?

Anyway, if the factory set up the bin for 12 PSI and I'm at 9, would that lower my fuel delivery proportionately, like 25% off? I need to dig up the equations someplace I guess.

What do you mean when you say "nice large displacement"? Is there significant variation among engines that are nominally 350ci? My motor is not factory built, but it's not especially beefed up either.

Last question while I've got you. Do you have any gut feeling for how much better VE engines get with the "breathing" opened up? I have an open element air cleaner, headers, no cat, big pipes, etc. Might that make a significant difference? I'd tend to think not at idle, since the throttle is probably the biggest restriction there...

Thanks again.

 

There are also numbers along the top flat of the injector. May need to wire brush a little to read them. If no paint then post those numbers.

GM Spec'd fuel pressure is 9 - 13 psi. Although your changing fuel pressure doesn't sound too good either. If the gauge is liquid filled they are inaccurate.

The 'nice large displacement' in the context I used it was in deference to 'too much displacement'. As there can never be too much displacement . I didn't know if maybe you were trying to feed a 383 or 400 or such with a 305 TBI unit.

Yes, that engine will breath better. As you mention thoguh it won't make that much of a difference at idle.

I'd start with the basics: fuel pressure and injector flow. See if you can get the numbers off of them.

RBob.

 

Here's the pressure gauge I'm using:

http://www.jegs.com/cgi-bin/ncommerc...97&prmenbr=361

I didn't realize that liquid filled was a bad thing. Have you seen them fluctuate, or just be off?

I'll try to figure out the injectors this afternoon...

Thanks.

 

Well, it turns out I have 5.0L injectors and I've been trying to use a stock 5.7L bin (on my 5.7L car). So no wonder it's lean. How much do you think that would put me off % wise? I see all kinds of conflicting numbers on the #/hour ratings of the injectors. The 5.0L range from 40-51 depending where I look, and the 5.7L range from 55-65 depending where I look.

If I used Grumpy's numbers, 40 and 55, then I'd be off by 38% on the lean side with these injectors.

Heck, that would put the 135 I started with up to 185 instead of 175 where I am now. Holy schmoly.

Things may be falling into place, but I'd sure love to have some flow numbers that people are confident in. I see different numbers all over.

Regarding fuel pressure gauges, I found several threads on this site that discussed how liquid filled pressure gauges fall apart at high temps because the liquid expands and creates counterpressure on the fuel - which lowers the reading. This is what I see, lower pressure at high temp. So I tend to think my pressure is good since I see 12 when the gauge is cold.

 

The larger the injector, the smaller the BPWc will be as its inversly proportional to the injector flow rate. The BPW of 135 sounds about right for the stock injectors. I remember mine being 135 as well.

 

But that was with the proper injectors, right? I mean, you have a 5.7L motor and the associated 5.7L injectors. I have a mismatch between the injectors, which are 5.0L, and my 5.7L motor. I'm sure I need to increase the BPW for this, I'm just not sure how much since I don't know the relative flow rates.

 

Increasing the BPC will work to get the engine running and drivable. However, the injectors will be out of pulse width and go static quickly.

Can either increase the fuel pressure or swap to larger injectors. To increase the fuel pressure will need a high pressure fuel pump. A TBI pump will not cut it.

If going to larger injectors try to get the Yellow/Brown 68 lb/hr ones.

The stock green/white 305 injectors are 55lb/hr. The orange/black 350 injectors are 61lb/hr, while the yellow/brown are used on the cop car 350's.

RBob.

 

If you'll take the time to teach me, I'd be interested to learn how I convert from BPC to actual base pulse width in ms. What VE and RPM would I have to get to in order to go static?

Can you recommend a good source for injectors? Junk yard, E-bay, something else?

Hmm. If those numbers are right, I'm only 10% off now. That wouldn't really explain what I'm seeing. It also doesn't seem like enough difference to bother changing them out. Is it?

 

Not easy. Bunch of conversions along with a bunch of factoring.

Best bet is to scour junkyards for an early BBC setup. That will give you bigger bores along with bigger injectors.


A stock truck 350 (61lb/hr) makes 200 HP (if that) and dies at 3800 RPM. The proper method of injector selection is to match it to the expected HP.

For an estimated injector size take the peak HP and divide by 4. Then multiply that answer by 1.2.

300 HP / 4 = 75
75 * 1.2 = 90lb/hr

If the heads are more efficient (chamber design) a smaller injector is required.

RBob.

 

Okay, I'll try and you tell me what's wrong.

Assume a 5.7L engine, 8 cylinders, so 0.7125 L/cylinder or 0.0007125 cu-m/cyl.

Assume the low end of estimates for 5L injectors, 40 pounds/hour. This is 18.14 kg/hr, 0.00504 kg/sec.

Assume we want 14.7 A/F (weight or mass) ratio.

Air is 1.225 kg/cu-m. Fuel, I think, is 6 lbs/gal, or 719 kg/cu-m.

Air volume plus fuel volume needs to be 0.007125 cu-m.
Air mass needs to be 14.7 times fuel mass.
Fuel mass is 719 times fuel volume.
Air mass is 1.225 times air volume.

Work it all out and I get fuel volume = 8.257e-8 cu-m.
Fuel mass is therefor 5.937e-5 kg (per cylinder firing).

At 0.00504 kg/sec, this takes 11.78 ms.

Invert this and we can stand 85 pulses/sec. But there are two injectors, so we can stand 170 pulses/sec.

170 pulses/sec is 10,187 sparks per minute, which I think is 5093 RPM (half of a revolution per spark).

So if everything is ideal, I get that we go static at 5100 RPM.

Now we actually probably want to be more rich than 14.7 at this point which will drop the RPM we can stand. The injectors also must have some wasted time while they are opening and closing, which will also make things worse.

If we go up to 12.7:1, then we'd drop to 4410 RPM, ignoring the injector thing. If we simplistically said that we want a 10% buffer for the injectors to open and close, that'd bring us around to 4000 RPM.

You said 3800 RPM, but I'm not sure what you were assuming for injector #/hr.

Anyway...

 

My mistake I thought you meant how the ECM calculates the PW. Only additional item to factor in is the amount of time available for the PW in use.

Here is what the ECM would come up with:

BPC: 175
CTS: 90°C
MAP: 100Kpa
AFR: 12.8
VE%: 100
BLM: 128
Bias: 369usec
RPM: 3200

9.372 msec, 99 DC% (8 cyl)

Note I set the BPC high such as with the 305 injectors. Here are the same parameters with a lower BPC:

BPC: 112
CTS: 90°C
MAP: 100Kpa
AFR: 12.8
VE%: 100
BLM: 128
Bias: 369usec
RPM: 4900

6.137 msec, 100 DC% (8 cyl)

This puts the RPM at 4900 before the injectors are static. And then at a lower BPC:

BPC: 80
CTS: 90°C
MAP: 100
AFR: 12.8
VE%: 100
BLM: 128
Bias: 369usec
RPM: 6000

4.489 msec, 89 DC% (8 cyl)

Note the 89% DC and RPM of 6,000. There is also a need for big injectors to supply that fuel. This isn't a good scenario either. Because the maximum PW will be a little after peak torque.

RBob.

 

Do you know how the ECM used the CTS reading? I assume it's trying to infer the intake air temp from the CTS? Seems like this could be a significant error source with an aftermarket intake manifold and air cleaner.

 

On the '7747 the ECM uses the CTS as the inverse temperature term for the PW calculation.

The intake air temperature is controlled by the use of a stove pipe heater on the aircleaner along with the water jacketed intake manifold (the whole thing is jacketed!).

Bypassing either or both of these intake air temperature control devices will throw the ECM calibration off. The stove pipe isn't as important to the calibration as the water jacketed intake manifold.

RBob.

 

I've got an open element air cleaner, so no stovepipe. I'm actually not sure whether I have the coolant flow up through the intake manifold. IIRC, the Edelbrock Performer can be set up with it, but some engine builders block off the passages with the gasket? Something like that.

Either way, since I'm adapting to a 4-bbl carb manifold, the air is probably cooler than expected.

What terms in a 7747 bin would I tweak on to make up for the non-factory air intake?